[gmx-users] How to concatenate tpr files?

[Justin Lemkul]

On 1/23/17 11:10 AM, leila karami wrote:
> Dear gromacs users,
>
> About the concatenation trajectory files, I have a question.
>
> I did 10 MD simulations (each section 10 ns). Now, I want to concatenate
> all of the trajectories and determine RMSD on whole trajectory. To do this,
> I used the following command:
>
> gmx trjcat -f md1.trr md2.trr md3.trr md4.trr md5.trr md6.trr md7.trr
> md8.trr md9.trr md10.trr -o all_md.xtc –settime
>
> The whole trajectory file made correctly. To determined RMSD, I used the
> following command.
>
> gmx  rms  -f  all_md.xtc  -s  *.tpr  -n index.ndx  -o rmsd.xvg  -tu  ns
>
> I want to know, which tpr file must be used? tpr1 or tpr2 or tpr3 or …?
>
> I tried this command with md1.tpr and md10.tpr and in each time, RMSD had
> the different values. So, should I concatenate the tpr files and use the
> complete tpr file.
>
> How can I concatenate the tpr files?
>

You don't.  A .tpr file has a configuration, velocities, state information, and 
a set of instructions.  The one you choose depends on how you created them.  If 
you re-invoked grompp to create them for each interval, they will each contain a 
different state, so naturally the result of an RMSD calculation will differ 
because there are different coordinates in each .tpr file.  If you use gmx 
convert-tpr to simply change the number of steps (much easier and cleaner!) then 
the coordinates are always the same, only the number of steps differs, so the 
.tpr files are not substantively different for the purposes of analysis.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================